Search

This review synthesizes research findings on the pain and welfare of dairy calves undergoing disbudding procedures. We describe disbudding practices in North America as well as the use and perceptions of pain control for these procedures. Governing bodies across Canada and the United States, including each country's veterinary medical association and nationwide initiatives such as proAction and Farmers Assuring Responsible Management (FARM), recommend or require the use of a local anesthetic, a nonsteroidal anti-inflammatory drug (NSAID), and a sedative for disbudding procedures. Although the use of pain relief for disbudding has increased over the past decade or so, some in the dairy industry still do not believe that pain control for disbudding is necessary. As a painful procedure, disbudding has numerous welfare impacts on the calf both during and following the procedure that can be categorized under all 3 principles of animal welfare: natural living, biological functioning, and affective state. The use of pain control for disbudding; namely, a local anesthetic and NSAID, can improve welfare outcomes such as procedure-induced pain behavior, cortisol concentrations, mechanical nociceptive threshold, emotional states, and so on, compared with no pain control for the procedure. Although extensive research exists on pain control practices for disbudding, this review identified further gaps in knowledge and areas for future research. Mechanical nociceptive threshold can be evaluated around the disbudding wounds and is a reliable test in older calves; however, this outcome in very young calves after caustic paste disbudding has been reported to be inconclusive compared with that in older calves. As well, research evaluating xylazine sedation for disbudding has reported both potentially positive and negative results that are difficult to interpret or base suggestions on for the use of this drug. Finally, wounds caused by disbudding take a long time to heal (up to 13 wk) and have increased sensitivity for the entire healing process. Therefore, future research should aim to (1) determine accurate behavioral tests for calves under 1 wk of age undergoing disbudding to better understand their experience, (2) further attempt to understand the effects of xylazine sedation for disbudding and potential impacts of providing this medication, and (3) determine more ways to reduce the healing time and pain experienced by the calf after disbudding procedures. [ABSTRACT FROM AUTHOR]

The objective of this study was to determine the effect of a biologically normal plane of nutrition compared with a limited plane on the primary outcome wound healing, and one dose of nonsteroidal anti-inflammatory drug (NSAID) compared with 2 on the secondary outcomes: lying behavior, haptoglobin concentrations, and mechanical nociceptive threshold (MNT) in calves disbudded via cautery iron. Eighty female Holstein calves were enrolled at birth, individually housed, and fed via a Calf Rail system (Förster Technik). A 2 × 2 factorial design was used to assess the effect of plane of nutrition and an additional NSAID. Calves were randomly assigned to a biologically normal plane of nutrition (BN; offered up to 15 L/d) or a limited plane (LP; offered up to 6 L/d) and to receive one or 2 doses of meloxicam. All calves received a lidocaine cornual nerve block and a subcutaneous injection of meloxicam 15 min before cautery disbudding at 18 to 25 d of age, and half the calves received an additional injection of meloxicam (0.5 mg/kg) 3 d after disbudding. Tissue type present, wound diameter, and wound depth were evaluated 2 times per week for 7 to 8 wk as measures of wound healing, lying behavior was recorded beginning 1 to 2 wk before disbudding until 7 to 8 wk after as a behavioral indicator of pain, haptoglobin concentrations were measured once per day for 7 d after disbudding, and MNT was evaluated 2 times/wk for 3 wk. Survival analyses were analyzed using Cox regression models (wound healing) and continuous data were analyzed using mixed-effect linear regression models. Only 12% of horn buds were completely healed by 7 to 8 wk after disbudding and 54% had re-epithelized at this time. At any time, wounds from BN calves were more likely to have had re-epithelization occur compared with wounds from LP calves (hazard ratio: 1.93, 95% CI: 1.18–3.14). Wounds from calves that received only one dose of NSAID were more likely to have re-epithelization occur, compared with wounds from calves given 2 doses (hazard ratio: 1.87, 95% CI: 1.15–3.05). Wounds from BN calves had smaller diameters and depths over time beginning on wk 3 compared with LP calves. Wounds from calves that received an additional NSAID had larger diameters and depths over time beginning on wk 4 and 3 respectively, compared with calves that only received one dose of NSAID. Calves that received an extra NSAID tended to be less sensitive 7, 10, and 17 d after disbudding compared with calves that only received one dose and spent less time lying in the week after disbudding. Calves on the BN milk program were more active compared with LP calves with lower lying times, fewer lying bouts per day, and longer average lying bouts. Our results indicate that a BN milk feeding program for calves can result in faster healing times and more activity, and that providing an extra NSAID 3 d after disbudding appears to slow the healing process but may result in less pain experienced by the calf 1 to 2 wk after the procedure. This study is also among the first to demonstrate that after the complete removal of the horn bud, wounds can take more than 8 weeks to re-epithelize and fully heal. [ABSTRACT FROM AUTHOR]

• RNNs are robust to many issues found when merging data from multiple dairy herds. • Poor Se/Sp with unmeasured variance means input variables must be chosen carefully. • Deep neural networks are robust to label noise seen in producer-recorded outcomes. • RNNs are robust to incomplete observations if remaining variables are informative. With growing adoption of precision dairy technologies, the use of big data is becoming increasingly common in the dairy industry. The speed at which data are generated has led to increased interest in developing detection and predictive models for animal health and disease events using real time records. When combining data from multiple sources, statistical methods exist to account for the underlying heterogeneity in data collected from commercial farms, although its impact on predictive models is not known. We investigated how 4 different issues commonly seen in these large datasets impact the performance of deep recurrent neural networks (RNNs) trained to detect the onset of clinical mastitis (CM) in dairy cows. Data were simulated by first sampling from real-world data and adding noise, then defining the association between predictor variables and CM while incorporating parameters to reflect underlying heterogeneity: 1) random effects to reflect unmeasured variability at the farm level (3 levels – none, moderate, high); 2) random effects to reflect unmeasured variability at the cow level (3 levels – none, moderate, high); 3) missed recording of CM cases (3 false-negative rates – 0.10, 0.25, 0.50); and 4) incomplete observations due to certain farms not having a somatic cell count sensor (SCC data missing vs SCC data included). At baseline (moderate farm and cow random effects; moderate misclassification; 42% herds with SCC sensor) the model achieved a sensitivity and specificity of 86% and 90% respectively. Higher levels of unmeasured variability at the farm and cow levels resulted in reduced model performance (sensitivity and specificity of 76% and 85% at the highest levels), indicating that data collection and feature selection should be informed by previous knowledge of the associations between the outcome and predictors when possible, and that model performance may be limited when predictors are selected only from routinely collected data. However, even when 50% of CM cases were incorrectly recorded as CM-negative, model performance did not decrease, demonstrating that deep RNNs are robust to the level of misclassification that would be typically encountered in dairy datasets. RNNs were also able to accurately detect CM onset even when a highly predictive variable, somatic cell count, was excluded from training and test data, but the models took longer to train. The effect of unmeasured variability on model performance demonstrates how predictors should be selected for RNNs, whereas RNNs appear to be very robust to misclassification in training data as well as missing variables. Researchers developing studies using deep learning should therefore focus their attention more on predictor selection than on reducing or standardizing outcome recording, since RNNs appear to be robust to the latter, while being more strongly impacted by the former. [ABSTRACT FROM AUTHOR]

• Recurrent neural networks can effectively identify dairy cows with mastitis. • Including additional variables improves detection over milk characteristics alone. • Within-day changes in behaviour and milk are key indicators of mastitis onset. • Between-farm variability is likely a major limiter for model generalizability. Mastitis is the most important disease in the dairy industry. With widespread adoption of automated milking systems (AMS) in Canada, there is an increasing need for automated detection of mastitis in AMS farms. The main objective of this study was to develop a recurrent neural network (RNN) model for the detection of clinical mastitis (CM) in dairy cows on farms using AMS. Producer-recorded treatment records and AMS data were collected over 3 time periods from a total of 89 dairy farms in 7 provinces across Canada. In addition to developing effective models for the detection of CM, our study also evaluated different windows around the day of diagnosis when the cow would be considered CM-positive to guide practical implementation of models. We also compared numerous subsets of variables including milk and behavioural characteristics, cow traits and farm-level/environmental variables to determine their importance and impact on model performance. Data were randomly divided into a training and a hold-out test set, consisting of all records from 66 and 23 farms, respectively. A 10-fold internal cross-validation was also employed on the training set for model development. When comparing different windows of time around diagnosis, considering animals as CM-positive for 3 d prior to recorded diagnosis resulted in the most timely and effective detection of CM with a per-case sensitivity of 89.8% (range: 83.3–96.0%), and per-day specificity of 84.3% (range: 83.4–85.8%) over the validation folds. These levels of sensitivity and specificity were achieved when using all recorded variables and their daily variances, although the inclusion of behavioural variables and farm-level/environmental variables provided marginal performance improvement over using milk traits alone. Performance of the model was worse on the hold-out test set with a per-case sensitivity of 83.5% (range: 77.9–86.3%) and a per-day specificity of 80.4 % (range: 78.1–82.4%), likely due to farm-specific heterogeneity not encountered in the training data. Over 90% of cases of severe CM (defined by an increase in milk temperature over the pre-CM baseline) were identified by the model, indicating effective performance for the detection of CM requiring the most immediate treatment. Somatic cell count, daily variance in milking interval and milk temperature were identified as the 3 most important variables defined by their impact on model predictions. In addition to milking traits, 8 of the top 20 variables were behavioural measurements, suggesting they can play a role in the detection of CM. Daily variances also represented 8 of the 20 most important variables indicating that CM onset may be associated with sudden, within-day changes in the animal. In conclusion, this study demonstrates that RNNs are able to effectively detect CM by integrating a number of variables that are regularly measured on AMS farms but have typically been excluded from CM detection models. [ABSTRACT FROM AUTHOR]